24 July 2021
AperTO - Archivio Istituzionale Open Access dell'Università di Torino
Original Citation:
Acoustic communication within ant societies and its mimicry by mutualistic and socially parasitic myrmecophiles
Published version:
DOI:10.1016/j.anbehav.2016.10.031 Terms of use:
Open Access
(Article begins on next page)
Anyone can freely access the full text of works made available as "Open Access". Works made available under a Creative Commons license can be used according to the terms and conditions of said license. Use of all other works requires consent of the right holder (author or publisher) if not exempted from copyright protection by the applicable law. Availability:
This is the author's manuscript
Page 1 of 31 1
2 3 4
This is an author version of the contribution published on:
5Questa è la versione dell’autore dell’opera:
6[Animal Behaviour,
online first
, 10.1016/j.anbehav.2016.10.031]
78
The definitive version is available at:
9La versione definitiva è disponibile alla URL:
10[http://www.sciencedirect.com/science/article/pii/S0003347216302822]
1112 13
Page 2 of 31
Acoustic communication within ant societies and its mimicry by
14mutualistic and socially parasitic myrmecophiles
1516
K Schönrogge1*, F. Barbero2, L.P. Casacci2, J Settele3, JA Thomas4
17 18 19 20 21
1. Centre for Ecology & Hydrology, MacLean Building, Benson Lane, Wallingford, OX10 22
8BB, UK 23
24
2. Department of Life Sciences and Systems Biology, University of Turin, Via Academia 25
Albertina, 10123, Turin, Italy 26
27
3. UFZ, Helmholtz Centre for Environmental Research, Department of Community Ecology, 28
Theodor-Lieser-Str. 4, 06120 Halle, Germany 29
30
4. Department of Zoology, University of Oxford, Oxford, South Parks Rd, OX1 3PS, United 31 Kingdom 32 33 34 35 36 37 38 39 40 41 * Corresponding author: 42
K. Schönrogge (ksc@ceh.ac.uk), Tel: +44 (0)1491 838800, Fax: +44 (0)1491 692424 43
44
Running title: Acoustic communication in ant – non-ant interactions 45
46
Word Count: 5390 (excl. references) 47
Page 3 of 31 Abstract:
49
This review focusses on the main acoustic adaptations that have evolved to enhance social 50
communication in ants. We also describe how other invertebrates mimic these acoustic 51
signals in order to coexist with ants in the case of mutualistic myrmecophiles, or, in the case 52
of social parasites, corrupt them in order to infiltrate ant societies and exploit their resources. 53
New data suggest that the strength of each ant-myrmecophile interaction leads to distinctive 54
sound profiles and may be a better predictor of the similarity of sound between different 55
myrmecophilous species than their phylogenetic distance. Finally, we discuss the 56
evolutionary significance of vibrations emitted by specialised myrmecophiles in the context of 57
ant multimodal communication involving the use of chemical and acoustic signals in 58
combination and identify future challenges for research including how new technology might 59
allow a yet better understanding of the study systems. 60
61 62
Keywords: Acoustic communication, ants, mutualists, social parasites, social structure 63
Page 4 of 31 Introduction
65
Efficient communication to coordinate the actions of up to a million specialised nestmates is 66
fundamental to the success of social insects, especially ants. Various modes of signalling 67
have been identified, including the release of semio-chemicals, visual behavioural displays 68
involving movement or posture, tactile interactions, and the comparatively poorly studied use 69
of acoustic signals (Hölldobler & Wilson, 1990, 2009). As hotspots of resources in their 70
environment, ants fiercely defend their colonies using a wide range of weapons (e.g. gland 71
secretions, mandibles, sting), which are deployed in the manner of co-ordinated attacks by 72
legions of intercommunicating workers. Nevertheless, ant nests are also magnets for other 73
organisms that have evolved means to overcome the hostility of the host ants. Thus, an 74
estimated ~10,000 invertebrate species live as obligate social parasites of ants, able to 75
penetrate and exploit the resources within host colonies in order to complete their life-cycle 76
(Thomas, Schönrogge, Elmes, 2005). The large majority of these adaptations evolved in 77
many separate lines, especially among Coleoptera, Diptera, Lepidoptera and other 78
Hymenoptera, from a ten-times greater number of commensals or mutualists (Fiedler, 1998; 79
Hölldobler & Wilson, 1990; Nash & Boomsma, 2008; Pierce et al., 2002; Thomas, 80
Schönrogge et al., 2005). All these myrmecophiles show morphological, behavioural, 81
chemical or acoustic adaptations to interact with ants (Cottrell, 1984; Donisthorpe, 1927; 82
Hinton, 1951; Lenoir, D'Ettorre, Errard, & Hefetz, 2001; Malicky, 1969; Wasmann, 1913; 83
Wheeler, 1910; Witek, Barbero, & Marko, 2014). Armour, stealth and the secretion of 84
attractive food rewards are frequently sufficient for unspecific or facultative myrmecophiles to 85
access the enemy-free spaces of ants. However, the subversion of the ants’ chemical and/or 86
acoustic signalling is generally required to enable true social parasites (sensu Nash & 87
Boomsma, 2008) to live for long periods as undetected intruders in close contact with their 88
hosts. 89
Page 5 of 31 A key element of successful co-habitation in ant nests is to circumvent the host’s ability to 90
differentiate between nestmates and intruders. Nestmate recognition is a dynamic process, 91
primarily based on the detection of distinctive species- or colony-specific cocktails of 92
cuticular hydrocarbons (CHC) covering the surface of all individuals (Hölldobler & Wilson, 93
1990; Howard, 1993; vander Meer & Morel, 1998; Winston, 1992). Social interactions such 94
as allogrooming ensure an exchange between the CHC mixtures among nestmates and give 95
rise to a shared CHC gestalt odour (vander Meer & Morel, 1998). The role that chemical 96
communication and nestmate recognition have in maintaining the cohesion of ant societies 97
and those of other social insects has been subject to extensive study, with excellent recent 98
reviews, for example by Martin & Drijfhout (2009) and van Wilgenburg, Symonds, & Elgar 99
(2011): The deployment of chemical communication by obligate social parasites to subvert 100
host recognition systems is equally well reviewed (e.g. Lenoir et al., 2001; von Thienen, 101
Metzler, Choe, & Witte, 2014). 102
In contrast, the function, the origin and role of acoustic signals in ants and their corruption by 103
social parasites are much less well studied. In this review, we therefore focus on the state of 104
the art concerning acoustic signaling in ants, and then consider the acoustic signaling of 105
obligate and facultative myrmecophiles. In both cases we emphasize the insights that have 106
resulted from recent technological advances that allow unalarmed ants and their guests to 107
be recorded and to receive broadcasts of their acoustic signals under semi-natural 108
conditions (Barbero, Thomas, et al., 2009; Riva, Barbero, Bonelli, Balletto, Casacci, in 109
press). 110
We first examine ant sound producing organs and convergent adaptations that allow non-ant 111
organisms to mimic and subvert ant–ant communications, focussing on advances in 112
knowledge since the reviews by Hölldobler & Wilson (1990), Fiedler (1998), Pierce and 113
colleagues (2002), Thomas and colleagues (2005) and Nash & Boomsma (2008), or covered 114
cursorily by Witek and colleagues (2014). We then review recent insights concerning the ant 115
acoustic signals themselves and their corruption by social parasites. This includes both the 116
Page 6 of 31 morphological adaptations to produce acoustic signals, the behavioural responses to them, 117
and thus the impact on ant – social parasite/guest interactions. Much of this builds on the 118
pioneering work of Markl (1965, 1967), DeVries (1991a, 1991b), Hölldobler, Braun, 119
Gronenberg, Kirchner, & Peeters (1994) and Kirchner (1997). Finally we present new data 120
relating the intimacy of interactions of lycaenid butterfly larvae to phylogeny and the similarity 121
of acoustic signalling. 122
Acoustic signalling in ants
123The use of acoustics, whether through receiving pressure waves through the air (i.e. sounds 124
stricto sensu) or substrate vibrations, is a common means of communication in insects, 125
whose functions include defence, displays of aggression, territorial signalling and mate 126
attraction (Bennet-Clark, 1998; Gerhardt & Huber, 2002). Its advantage as a signal over 127
chemical volatiles lies in instantaneous reception that pinpoints a distant, but exact, location 128
to the receiver, for example in social insects to attract help (Markl, 1965, 1967; Roces, 129
Tautz, & Hölldobler, 1993). The physics, use and effects of substrate-borne vibrations of 130
ants and other insects are comprehensively reviewed by P.S. Hill (2009). A simple form 131
involves “drumming”, where the substrate is tapped by part of the exoskeleton to produce 132
vibrations. Drumming is employed by many ant taxa, but at least four of the eleven 133
subfamilies also stridulate by rasping a ‘plectrum’ across a ‘file’ (pars stridens), both 134
chitinous organs being located on opposite segments of the anterior abdomen (see Fig. 1 k-135
o, u-y) (Barbero, Thomas, Bonelli, Balletto, & Schönrogge, 2009b; Golden & P.S. Hill, 2016; 136
Ruiz, Martinez, Martinez, & Hernandez, 2006). Although these stridulations produce air-137
borne (as well as substrate-borne) pressure waves that are audible to the human ear, it 138
remains uncertain whether ants can perceive sound as pressure waves through the air 139
(Hickling & Brown, 2000, 2001; Roces & Tautz, 2001). In contrast, there is no controversy 140
about the ants’ ability to perceive substrate vibrations and two types of sensor have been 141
proposed to receive substrate vibrations: campaniform sensilla measuring the tension in the 142
Page 7 of 31 exoskeleton; and the subgenual organ, a spherical arrangement of sensory cells in the tibia, 143
as described from Camponotus ligniperda (Gronenberg, 1996; Menzel & Tautz, 1994). 144
Most studies that measure insect acoustics have used accelerometers, moving coil- or 145
particle velocity microphones, often with phase inversion focussing on the vibrational part of 146
the signal rather than pressure waves through the air. Hereafter in this review we use the 147
term “sound” sensu latu in its broadest sense, as we do the terms: calls, vibrations, vibro-148
acoustics and stridulations. 149
Early studies suggested that acoustic signals were a minor means of communication among 150
ants, largely confined to activities outside the nest and mainly signalling alarm or calls for 151
rescue, for instance when parts of nests collapse (Markl 1965, 1967). Due to a perceived 152
preponderance of stridulatory organs among soil nesting ant species, Markl (1973) 153
hypothesised that stridulation evolved initially as a burial/rescue signal when volatile 154
chemicals would be ineffective, whereas substrate borne vibrations would at least travel 155
short distances. However, this is not supported by Golden and P.S. Hill (2016), who showed 156
that stridulation organs have evolved independently multiple times in ants. In addition, 157
whereas Markl (1973) suggested that they would probably become vestigial over time in 158
arboreal ant species, due to the rarity of burial by soil, there was instead a strong positive 159
association between the presence of functional stridulation organs and the possession of an 160
arboreal life–style (Golden & P.S. Hill, 2016). 161
Nestmate recruitment is the most frequently reported function for ant–ant acoustic signalling. 162
For example, outside the nest, Atta cephalotes uses vibratory signals to attract foraging 163
workers towards newly found food sources (Roces & Hölldobler 1995). The same authors 164
also observed that in the presence of parasitic phorid flies, foragers used acoustics to recruit 165
minor workers for defence, thus also employing vibrations as alarm signals (Roces & 166
Hölldobler, 1995, 1996). Finally, although created by a scraper and file organ located on the 167
first gastric tergite and the post-petiole, Tautz and colleagues (1995) observed that 168
Page 8 of 31 vibrations travelled the length of the body to the mandibles, aiding the cutting of soft young 169
leaf tissue by stiffening it. Behavioural experiments, however, suggest that this is a 170
secondary effect and that communication is the main function for these vibrations (Roces & 171
Hölldobler, 1996). 172
It has recently become clear that acoustic signals are also used to transmit more abstract 173
information, including a species’ identity or an individual’s caste and status (Barbero, 174
Thomas et al., 2009; Casacci et al., 2013; Ferreira, Cros, Fresneau, & Rybak, 2014). For 175
example, modern molecular analyses revealed the neotropical ponerine ant species, 176
Pachycondyla apicalis, to be a species complex of five cryptic lineages. The stridulations of 177
three largely sympatric lineages are also distinctive, suggesting that morphological 178
characters on the pars stridens differ in length, width and ridge gap in each lineage (Ferreira, 179
Cros, Fresneau, & Rybak, 2014; Wild, 2005). By contrast, two allopatric lineages had very 180
similar acoustics, suggesting disruptive selection on this trait where sympatric overlap is 181
high. 182
Acoustic patterns also signal caste and hierarchical status in at least two genera of 183
Myrmicinae ants: Myrmica (Barbero, Thomas et al., 2009) and Pheidole (Di Giulio et al., 184
2015). In both taxa, the queens produce distinctive stridulations which, when played back to 185
kin workers, elicit additional ‘royal’ protective behaviours compared with responses to worker 186
signals (Barbero, Bonelli, Thomas, Balletto, & Schönrogge, 2009; Barbero & Casacci, 2015; 187
Barbero, Thomas et al., 2009; Casacci et al., 2013; Ferreira, Poteaux, Delabie, Fresneau, & 188
Rybak, 2010). In addition, in Pheidole pallidula the soldier and minor worker castes also 189
make distinctive vibroacoustic signals (Di Giulio et al., 2015). Unlike Pachycondyla species, 190
little inter-specific variation was detected in either the queen- or worker-sounds made by 191
closely-related sympatric species of Myrmica (Barbero et al., 2012; Barbero, Thomas et al., 192
2009; Thomas, Schönrogge, Bonelli, Barbero, & Balletto, 2010), which are instead clearly 193
demarcated by unique hydrocarbon profiles (Elmes, Akino, Thomas, Clarke, & Knapp, 194
Page 9 of 31 2002). Although the young stages of tested ants are mute (e.g. DeVries, Cocroft, & Thomas, 195
1993), Casacci and colleagues (2013) found that acoustic signalling appears to act as a 196
substitute for other forms of communication in developing Myrmica pupae. The various 197
stages of ant brood, from egg to pupa, are afforded ascending levels of priority based on 198
tactile and chemical cues (Brian, 1975). Most are mute, but the older “brown”, sclerotised 199
pupae of Myrmica species produce calls, emitted as single pulses, similar to those of 200
workers (Casacci et al. 2013). This coincides with a presumed reduced ability to secrete 201
brood recognition pheromones during this period, and brown pupae that were experimentally 202
silenced fell significantly behind their mute white siblings in social standing. 203
Acoustic signals of myrmecophiles
204Derived acoustic signals that enhance interactions with ants are increasingly being 205
confirmed in both juvenile and adult stages of myrmecophiles. To date, most studies involve 206
riodinid and, especially, lycaenid butterfly larvae and pupae (e.g. Barbero, Thomas et al., 207
2009; DeVries, 1990, 1991a; Pierce et al., 2002). However, similar phenomena were 208
recently described from adults of a socially parasitic beetle, Paussus favieri (Di Giulio et al., 209
2015), where males and females emit mimetic stridulations using a row of scrapers on the 210
proximal abdominal segment rasping across a file located on the hind femora (see Fig. 1p-t). 211
Stridulation organs 212
With a few exceptions, an ability to produce calls occurs after the third larval moult in riodinid 213
and lycaenid larvae, coinciding with the development of chemical ‘ant organs’, which 214
perhaps suggests they act synergistically (DeVries, 1991a). In most riodinids, acoustic 215
signals are generated by grooved vibratory papillae. These are typically found in pairs on the 216
prothorax, and grate against specialised epicranial granulations when the larva rotates its 217
head (see Fig 1a-e), especially when walking or under attack, generating low amplitude 218
substrate-borne calls (DeVries, 1991a). The tribe Eurybiini lacks vibratory papillae; instead, 219
Page 10 of 31 caterpillars generate calls by scraping teeth on a prothoracic cervical membrane against the 220
epicranial granulations in at least some mutualists or entomophagous predators of ant-221
tended Homoptera (DeVries & Penz, 2002; Travassos, DeVries, & Pierce, 2008). The 222
detection of dedicated organs in lycaenid larvae that produce calls has been elusive, apart 223
from a file-and-scraper described between the 5th and 6th abdominal segments of Arhopala
224
madytus (C. J. Hill, 1993) and a putative organ in Maculinea rebeli larvae (see Fig.1fg). In 225
other species strong substrate-borne vibrations (and apparently weak air-borne sounds) may 226
be generated by muscular contractions of the abdomen, which compress air through the 227
tracheae to produce distinctive rhythms and intensities in the manner of a wind instrument, 228
as described by Schurian and Fiedler (1991) for Polyommatus dezinus. These vibroacoustic 229
signals range from low background calls punctuated by pulses in mutualists (DeVries, 230
1991a) to the grunts, drumming and hisses of the host-specific Jalmenus evagoras 231
(Travassos & Pierce, 2000), to the mimetic calls of Maculinea larvae (Barbero, Bonelli et al., 232
2009; DeVries et al., 1993; Sala, Casacci, Balletto, Bonelli, & Barbero, 2014). 233
In contrast, the pupae of all lycaenids studied (Pierce et al., 2002) and a minority of riodinids 234
(DeVries, 1991a; Downey & Allyn, 1973; 1978; Ross, 1966) have a well-developed file-and-235
scraper organ (two pairs in the case of riodinids) situated between opposite segments of the 236
abdomen, that emit substrate- and air-borne calls often audible to humans (see Fig 1h-j). In 237
lycaenids, the plate against which teeth are rubbed may be complex, consisting of tubercles, 238
reticulations or ridges (Alvarez, Munguira, & Martinez-Ibanez, 2014). 239
Acoustic signalling in ant–myrmecophile interactions 240
Evidence that the acoustics of myrmecophiles are adaptive to their interactions with ants has 241
progressed from correlative studies to two experimental approaches: muting the 242
myrmecophile or recording and playing back their calls to undisturbed ant colonies. 243
First, DeVries (1991c) showed that fewer ants attended larvae of the mutualistic riodinid 244
Thisbe irenea that had been artificially silenced compared with controls that were able to 245
Page 11 of 31 call, establishing that at least one function of riodinid calls is to attract ants. Similarly, 246
Travassos and Pierce (2000) demonstrated that pupae of the lycaenid Jalmenus evagoras 247
stridulated more frequently in the presence of Iridomyrmex anceps ants, and attracted and 248
maintained a larger number of guards than muted ones. The calls convey the pupa’s value 249
as a provider of nutritious secretions to the ants, which does however, represent a significant 250
cost to the pupae. Tended pupae have been shown to lose 25% of weight and take longer to 251
eclose than untended ones (Pierce, Kitching, Buckley, Taylor, & Benbow, 1987). In further 252
behavioural experiments Travassos and Pierce (2000) showed that pupae used acoustic 253
signalling to adjust the number of attendant ants. They provided a path from an I. anceps 254
nest to signalling pupae and scored the rate of worker movement in relation to signal 255
strength once the pupa was discovered. This appears to be an important fitness component 256
evolved to attract no more than an adequate number of ant guards against enemy attacks. 257
The larvae of J. evagoras produce more varied acoustic signals than pupae - grunts, hisses 258
and drumming – and are also heavily attended and guarded by their mutualist ant (Pierce et 259
al., 2002). Hisses are emitted briefly after encountering a worker, whereas grunts are 260
produced throughout ant attendance. The ability of J. evagoras juveniles to produce distinct 261
vibrations, some probably with different functions, suggests the evolution of a finely-tuned 262
acoustic system of communication with their hosts, which might be elucidated using play-263
back experiments. 264
In parasitic interactions with ant colonies, the clearest evidence to date that some acoustic 265
signals are mimetic involves the highly specialized species of the Myrmica ant - Maculinea 266
butterfly and Pheidole ant - Paussus beetle systems. Initially, DeVries and colleagues (1993) 267
showed that the calls made by larvae of four Maculinea species differed from those of 268
phytophagous lycaenids in showing distinctive pulses that resembled the stridulations of 269
Myrmica worker ants. This was the first suggestion of mimicry of an adult host attribute by 270
the caterpillars, which appeared to be genus- rather than species-specific. The insects in 271
early experiments were unavoidably alarmed, being held with forceps during the recording, 272
Page 12 of 31 but a similar genus-specific result was later obtained using modern equipment and 273
unstressed ants and butterflies. Both the pupae and larvae of Maculinea species closely 274
mimicked three attributes of their hosts’ acoustic signals: dominant frequency, pulse length, 275
pulse repetition frequency (Barbero, Bonelli et al., 2009, Barbero, Thomas et al., 2009). 276
However, the calls of both stages were significantly more similar to queen ant calls than they 277
were to worker calls, despite each being generated in a different way (see Fig.1f-j). 278
Behavioural bioassays, where the calls of butterflies and ants were played back to 279
unstressed Myrmica workers, revealed that the calls of juvenile Maculinea, especially those 280
of pupae, caused workers to respond as they do to queen ant calls. Both types of acoustic 281
stimuli caused worker ants to aggregate, antennate the source of sound, and show 282
significantly higher levels of guarding behaviour than was elicited in response to worker ant 283
calls (Barbero, Thomas et al., 2009). 284
Similar, but more sophisticated communication, was recently described between the carabid 285
beetle Paussus favieri, an obligate social parasite in all stages of its life-cycle, and their host 286
ant Pheidole pallidula (Di Giulio et al., 2015). Here the adult beetle can generate three types 287
of call when it stridulates, which respectively mimic the calls made by the queens, the 288
soldiers and the minor worker caste of its host. These calls elicit a range of responses when 289
played back to worker ants, consistent with the intruder’s more diverse activities (compared 290
to juvenile Maculinea) in different parts of the host’s society and nest. Thus P. favieri’s 291
various stridulations can elicit recruitment, including digging (rescue) behaviour, as well as 292
the enhanced level of ‘royal’ (queen ant) protection observed towards Maculinea pupae and 293
larvae. 294
[insert Figure 1] 295
Larval acoustic signals and phylogeny in the Lycaenidae 296
Various authors (e.g. DeVries, 1991a, 1991b; Fiedler, 1998; Pech, Fric, Konvicka, & Zrzavy, 297
2004; Pellissier, Litsios, Guisan, & Alvarez, 2012; Pierce et al., 2002) have analysed the 298
Page 13 of 31 evolution of myrmecophily in lycaenids and riodinids, including social parasitism in the 299
Lycaenidae, and most concluded that it also provided a template for diversification and 300
radiation in these species-rich families. Pierce and colleagues (2002) argued convincingly 301
that social parasitism (including entomophagy of the domestic Hemiptera of ants) has 302
evolved independently in at least 20 lineages. 303
The analysis of acoustics as a parameter in evolutionary studies of these taxa was 304
pioneered by DeVries (1991a, 1991b). In seminal early papers, DeVries (1991a, 1991b) 305
found that only lycaenids and riodinids that interacted with ants produced calls, while several 306
non myrmecophilous members of the tribe Eumaeini were silent. Subsequent studies and 307
reviews confirmed this pattern (e.g. Fiedler, Seufert, Maschwitz, & Idris, 1995) and provided 308
evidence of the use of lycaenid calls in enhancing the interaction with ants (Pierce et al., 309
2002; Barbero, Thomas et al., 2009, Sala et al. 2014). However, some lycaenid and riodinid 310
larvae and pupae also emit sounds when disturbed by putative predators or parasitoids, 311
even if ants are absent. In addition, other species classed as having no interaction with ants 312
do emit sound (e.g. Alvarez et al., 2014; Downey & Allyn, 1973; 1978; Fiedler, 1992, 1994; 313
Schurian & Fiedler, 1991). The most recent study, by Riva and colleagues (in press), found 314
that lycaenid sounds are highly specific and are emitted by both non- and myrmecophilous 315
species. Calls by species that are least associated with ants consist of shorter and more 316
distant pulses relative to those of species that are highly dependent on them. 317
Here we further explore the hypothesis that the strength of ant-myrmecophile interactions 318
(using Fiedler’s 1991 definitions) leads to characteristic sound profiles that may be a better 319
predictor of the similarity of sound between species than their phylogenetic distance. We 320
present a new analysis of the acoustic profiles made by 13 species of European lycaenids, 321
ranging from highly integrated ‘cuckoo’ social parasites (Maculinea alcon, Ma. rebeli) via one 322
host-specific mutualist (Plebejus argus) and a spectrum of generalist myrmecophiles, to 323
species for which little or no interaction is known (Lycaena spp.). The 13 species (see Fig. 2) 324
are a subset of the commensal or mutualistic species used by Riva and colleagues (in 325
Page 14 of 31 press), with three species of Maculinea added to represent the two levels of intimate 326
integration found in this socially parasitic genus (Thomas, Schönrogge et al., 2005). 327
Fourth instar caterpillars were recorded using customized equipment, as described by Riva 328
and colleagues (in press). We analyzed recordings of three individuals per species, 329
randomly selecting two trains of five pulses in each trace. Fourteen sound parameters were 330
measured using Praat v. 5.3.53 (Boersma & Weenink, 2013). These included the lower and 331
higher quartiles of the energy spectrum (Hz), power (dB2), intensity (dB), the
root-mean-332
square intensity level (dB) and the relation of the frequency peak energy to the call total 333
energy (%). Two temporal variables were measured from the oscillogram: the duration of the 334
pulse (s) and the Pulse Rate (calculated as 1/tstart(x) - tstart(x+1); s-1). Six additional variables
335
were estimated on each pulse by inspection of power spectra: the frequency of the first, 336
second and third peak amplitudes (Hz), the intensity of the first two peaks (dB) and the 337
center of gravity (Hz). 338
Hierarchical Cluster analyses was performed on a matrix of normalized Euclidean distances 339
over sound parameters, averaged by individual using unweighted pair-group average 340
(UPGMA) in Primer v. 6.1.12 (Primer-E Ltd.). A two-sample t - test was used to compare 341
differences between group distances.To test whether species differences reflect degrees of 342
myrmecophily, we used Phylogenetic Regression as implemented in the library “phyreg” 343
(Grafen, 1989) using R (R Core Team, 2015). Principal components, derived by PCA on log-344
transformed sound parameters, were correlated with the degree of myrmecophily while 345
controlling for phylogenetic relatedness among species. To assemble a working phylogeny, 346
we used cytochrome oxidase subunit 1 (COI) sequences of the 13 lycaenid species from two 347
recent studies on the Romanian and Iberian butterflies (Dinca et al., 2015; Dinca, Zakharov, 348
Hebert, & Vila, 2011). Geneious Pro 4.7.5 (Biomatters, http://www.geneious.com/) was used 349
to align COI sequences and to produce a neighbor-joining (NJ) tree. We also included in the 350
phylogeny Hamearis lucina (Riodinidae) and Pieris rapae (Pieridae) as outgroups. 351
Page 15 of 31 Two trees for species’ phylogenetic distance and for the similarity of acoustic profiles are 352
presented in Figure 2, together with the score for myrmecophily of each species. Similarities 353
in sound profiles neatly match the spectrum of observed strengths and specificities in 354
myrmecophily across the study species, much more closely than does phylogeny. Overall, 355
PC1 of the acoustic parameters explained 56% and PC2 a further 27% of variation, and both 356
were significantly correlated with the differences in myrmecophilous relationships (PC1: F1,13
357
= 11.146, P = 0.005; PC2: F1,13 = 6.959, P = 0.020) after accounting for phylogeny using
358
phylogenetic regression. 359
It is apparent that the sound profiles of Ma. rebeli and Ma. alcon (average Euclidean 360
distance (± 1SD) between Ma. rebeli and Ma. alcon = 1.65 ± 0.14) are far removed from all 361
other species, including from their congeners Ma. arion and Ma. teleius (Barbero, Bonelli et 362
al., 2009; Sala et al., 2014). Indeed, the mean Euclidean distances in the acoustic signals of 363
Ma. alcon or Ma. rebeli from other lycaenid species are among the highest measured to date 364
(mean Euclidean acoustic distance of Ma. alcon vs. lycaenids other than M. rebeli: 7.41 ± 365
1.00; Ma. rebeli vs lycaenids other than Ma. alcon: 7.66 ± 1.01; see also Riva et al. in press). 366
This is consistent with the intimate level of social integration these species achieve within 367
host ant nests, an association that is so close that in times of shortage the ants kill their own 368
brood to feed to these ‘cuckoos’ in the nest (Thomas, Elmes, Schönrogge, Simcox, & 369
Settele, 2005). It is also notable that the acoustics of Plebejus argus, the only host-specific 370
myrmecophile among the mutualistic species, is less similar to its nearest relative Plebejus 371
argyrognomon, and appears to converge with the two ‘predatory’ Maculinea social parasites 372
even though its ‘host’ ant, Lasius niger, has no known stridulation organs and belongs to a 373
different subfamily to Myrmica (mean Euclidean acoustic distance of P. argus vs. P. 374
argyrognomon: 4.33 ± 0.30; P. argus vs M. arion: 2.51 ± 0.55; paired t test: t16 = -8.723, P <
375
0.001; distance of P. argus vs. Ma. teleius: 3.79 ± 0.28; paired t test: t16 = -3.963, P = 0.001).
376
Scolitantides orion perhaps represents selection in the opposite direction to P. argus, being 377
less host specific than its ancestry or relatives might suggest, as, less convincingly, may 378
Page 16 of 31 Polyommatus icarus. Yet despite L. coridon and L. bellargus being close congeners, sounds 379
emitted by L. bellargus are much more similar to those produced by P. argyrognomon 380
(belonging to the same myrmecophilous category - 3) rather than to L. coridon (mean 381
Euclidean acoustic distance of L. coridon vs L. bellargus: 3.87 ± 0.15; P. argyrognomon vs L. 382
bellargus: 1.54 ± 0.20; paired t test: t16 = 27.775, P < 0.001). A possible, but untested,
383
explanation is that this reflects a similar disruptive selection via acoustics to that described in 384
sympatric lineages of the ant Pachycondyla, since the juveniles of these congeneric 385
butterflies overlap largely in distribution, sharing the same single species of foodplant and 386
often the same individual plant. 387
However, given the small number of species studied, we caution against over-interpreting 388
the apparent patterns depicted in Figure 2, and suggest they be tested by comparative 389
behavioural experimentation. We also recognise that vibrations of less- or non-390
myrmecophilous lycaenids (and other taxa) may have very different functions, such as 391
repelling natural enemies (Bura, Fleming, & Yack, 2009; Bura, Rohwer, Martin, & Yack, 392
2011). We tentatively suggest that ancestral species in the Lycaenidae were preadapted to 393
myrmecophily through an ability to make sounds, and that once behavioural relationships 394
with ants evolved, the selection regime changed resulting in adaptive mimetic sound profiles, 395
at least among obligate myrmecophiles. 396
[insert Figure 2] 397
Conclusions & Future Research
398Ants are known to sometimes use multiple cues to moderate kin behaviour, for example by 399
combining posturing, tactile and chemical interactions to convey complex or sequential 400
information and to elicit particular responses between members of their society (Hölldobler & 401
Wilson, 1990). To date little is known of how acoustic signalling might interact with other 402
Page 17 of 31 means of communication, and less still of whether myrmecophiles manipulate behaviour 403
using multiple cues. 404
Sound may be used synergistically with other modes of signalling. Hölldobler and colleagues 405
(1994) studied the role of audible vibrational signals made by the Ponerine ant Megaponera 406
foetens, a raider of termite colonies, in the context of trail following and column building.
407
They found that stridulations were emitted only during disturbances and for predator 408
avoidance. It is also known that M. foetens has a distinctive pheromone to signal alarm 409
(Janssen, Bestmann, Hölldobler, & Kern, 1995). These observations suggest that vibrations 410
may be used to qualify a general alarm signal that is chemical, but again this requires formal 411
testing. This is in contrast to the observations by Casacci and colleagues (2013) described 412
above where acoustic signalling appears to replace chemical and tactile signal apparently 413
with the same function of signalling rank, but this is not truly a case of multimodal 414
communication. 415
To date, no direct evidence exists for the behavioural consequences of full synergistic 416
multimodal communication involving acoustics. Yet the interactions of Maculinea butterfly 417
larvae and their Myrmica host ant societies illustrate the importance of both chemical and 418
acoustic mimicry. Here, the acceptance (or rejection) of larvae as members of their host 419
colony appears to be based entirely on a mimetic mixture of chemical secretions, but on this 420
cue alone intruders are treated simply like the low-ranking kin brood (Akino, Knapp, Thomas, 421
& Elmes, 1999; Thomas et al., 2013; Thomas, Schönrogge et al., 2005). It is the ability 422
simultaneously to emit acoustic calls that mimic adult hosts, and furthermore mimic queen 423
sounds, that is believed to explain the observed priority ‘royal’ behaviour that workers 424
regularly afford to social parasites, giving them a status that exceeds that of large ant larvae. 425
Not only do these brood parasites gain priority in the distribution of food by nursery workers 426
to the extent that workers feed younger kin ant brood to the Maculinea larvae when food is 427
short, but they are also carried ahead of kin ant brood when moving nest or during rescues 428
Page 18 of 31 (Elmes, 1989; Gerrish, 1994; Thomas, Schönrogge, et al., 2005). Anecdotal observations of 429
the manipulation of Paussus favieri by the beetle Pheidole pallidula suggests a similar 430
chemical-acoustic mechanism (Di Giulio et al., 2015), but as with ant-ant communication 431
itself, the putative use of acoustics in multimodal communication requires rigorous testing. 432
About 10,000 species of invertebrates from 11 orders are estimated have evolved 433
adaptations to infiltrate ant societies and live as parasites inside nests (Hölldobler & Wilson, 434
1990). Current studies have largely focussed on the family Lycaenidae among the 435
Lepidoptera and a few selected species of Coleoptera. While the study systems used today 436
provide some variety in the type of interactions with their host ants, there is clearly a vast 437
variety still to be discovered to understand respective roles of signalling modes and the 438
social interactions in ants and other social insects. 439
The important role that acoustic signalling has in ant- and other social insect societies is well 440
established and it is perhaps unsurprising that other, interacting species show adaptations 441
that relate to the hosts acoustic traits. In only a few cases, however, has the role of vibro-442
acoustics in mediating myrmecophile - host interactions been investigated experimentally. 443
The modalities of signal production, transmission and reception remain largely unknown for 444
most species of myrmecophiles or indeed their hosts, but the greatest future challenge is to 445
understand how different modes of signalling interact. Social insects are well known to 446
interpret stimuli in a context-dependent manner, where the same stimulus can trigger a 447
different behaviour when encountered under different circumstances (Hölldobler & Wilson 448
1990). Other aspects of insect social behaviour have been subject to sophisticated and 449
successful experimentation, and it should be possible to unravel this essential aspect of 450
communication. Hunt and Richards (2013) suggested that understanding the suites of 451
modalities in signalling enables a clearer view of the adaptive role of multimodal 452
communication, and while that has been true for rare examples such as the honey bee 453
waggle dance, research into understanding the role of ant acoustics is in its infancy. With the 454
development of recording equipment that is portable, affordable, which can focus on 455
Page 19 of 31 individuals and record sound and behaviour at the same time, our understanding of social 456
interactions should become more specific. Such instruments, laser-vibrometers and hand-457
held “noses” for acoustic and chemical analyses, are being developed for engineering 458
applications and could be deployed to record acoustic and chemical signals in behavioural 459
science in the near future. Technological developments in both recording equipment and 460
behavioural experimentation will allow designing studies following the same principles to 461
investigate synergistic effects of multiple chemical signals. 462
Acknowledgments 463
The authors collaborated within the project CLIMIT (Climate Change Impacts on Insects and 464
their Mitigation), funded by Deutsches Zentrum für Luft-und Raumfahrt-Bundesministerium 465
für Bildung und Forschung (Germany); Natural Environment Research Council (NERC) and 466
Department for Environment, Food, and Rural Affairs (UK); Agence nationale de la 467
recherche (France); Formas (Sweden); and Swedish Environmental Protection Agency 468
(Sweden) through the FP6 BiodivERsA Eranet. Part of the research was funded by the 469
Italian Ministry of Education, University, and Research (MIUR) within the project ‘‘A multitaxa 470
approach to study the impact of climate change on the biodiversity of Italian ecosystems.’ 471
Ethical Note 472
The authors confirm that their work adheres to the ASAB/ABS and ARRIVE Guidelines. The 473
guide to ethical information required for papers published in the journal has been consulted 474
as well. Maculinea caterpillars were collected under permit from The Italian Ministry for the 475
Environment (protocol numbers: 446/05. DPN/2D/2005/13993 & 0012494/PNM/2015). 476
The authors declare that there is no conflict of interest. 477
References
478
Akino, T., Knapp, J. J., Thomas, J. A., & Elmes, G. W. (1999). Chemical mimicry and host 479
specificity in the butterfly Maculinea rebeli, a social parasite of Myrmica ant colonies. 480
Page 20 of 31 Proceedings of the Royal Society of London Series B-Biological Sciences, 266, 481
1419-1426. 482
Alvarez, M., Munguira, M. L., & Martinez-Ibanez, M. D. (2014). Comparative study of the 483
morphology of stridulatory organs of the Iberian lycaenid butterfly pupae 484
(Lepidoptera). Journal of Morphology, 275, 414-430. 485
Barbero, F., Bonelli, S., Thomas, J. A., Balletto, E., & Schönrogge, K. (2009a). Acoustical 486
mimicry in a predatory social parasite of ants. Journal of Experimental Biology, 212, 487
4084-4090. 488
Barbero, F., & Casacci, L. P. (2015). Butterflies that trick ants with sound. Physics Today, 489
68, 64-65. 490
Barbero, F., Patricelli, D., Witek, M., Balletto, E., Casacci, L. P., Sala, M., & Bonelli, S. 491
(2012). Myrmica ants and their butterfly parasites with special focus on the acoustic 492
communication. Psyche: A Journal of Entomology, 2012, 1-11. 493
Barbero, F., Thomas, J. A., Bonelli, S., Balletto, E., & Schönrogge, K. (2009). Queen ants 494
make distinctive sounds that are mimicked by a butterfly social parasite. Science, 495
323, 782-785. 496
Bennet-Clark, H. C. (1998). Size and scale effects as constraints in insect sound 497
communication. Philosophical Transactions of the Royal Society B, 353, 407 - 419. 498
Boersma, P., & Weenink, D. (2013). Praat: Doing phonetics by computer [Computer 499
program]. Version 5.3.53. (2013) retrieved from http://www.praat.org. 500
Brian, M. V. (1975). Larval recognition by workers of the ant Myrmica. Animal Behaviour, 23, 501
745 - 756. 502
Bura, V. L., Fleming, A. J., & Yack, J. E. (2009). What's the buzz? Ultrasonic and sonic 503
warning signals in caterpillars of the great peacock moth (Saturnia pyri). 504
Naturwissenschaften, 96, 713-718. 505
Bura, V. L., Rohwer, V. G., Martin, P. R., & Yack, J. E. (2011). Whistling in caterpillars 506
(Amorpha juglandis, Bombycoidea): sound-producing mechanism and function. The 507
Journal of Experimental Biology, 214, 30-37. 508
Page 21 of 31 Casacci, L. P., Thomas, J. A., Sala, M., Treanor, D., Bonelli, S., Balletto, E., & Schönrogge, 509
K. (2013). Ant Pupae Employ Acoustics to Communicate Social Status in Their 510
Colony’s Hierarchy. Current Biology, 23, 323-327. 511
Cottrell, C. B. (1984). Aphytophagy in Butterflies - Its Relationship to Myrmecophily. 512
Zoological Journal of the Linnean Society, 80, 1-57. 513
DeVries, P. J. (1990). Enhancement of symbioses between butterfly caterpillars and ants by 514
vibrational communication. Science, 248, 1104-1106. 515
DeVries, P. J. (1991a). Call production by myrmecophilous riodinid and lycaenid butterfly 516
caterpillars (Lepidoptera): morphological, acoustical, functional, and evolutionary 517
patterns. American Museum Novitates, 3025, 1-23. 518
DeVries, P. J. (1991b). Evolutionary and Ecological Patterns in Myrmecophilous-Riodinid 519
Butterflies. In C. R. Huxley & D. F. Cutler (Eds.), Ant - Plant Interactions (pp. 143-520
156). Oxford: Oxford University Press. 521
DeVries, P. J. (1991c). Mutualism between Thisbe irenae butterflies and ants, and the role of 522
ant ecology in the evolution of larval - ant associations. Biological Journal of the 523
Linnean Society, 43, 179 - 195. 524
DeVries, P. J., Cocroft, R. B., & Thomas, J. A. (1993). Comparison of acoustical signals in 525
Maculinea butterfly caterpillars and their obligate host Myrmica ants. Biological 526
Journal of the Linnean Society, 49, 229-238. 527
DeVries, P. J., & Penz, C. M. (2002). Early stages of the entomophagous metelmark 528
butterfly Alesa amesis (Riodinidae: Eurybiini). Journal of the Lepidopterist's Society, 529
56, 265 - 271. 530
Di Giulio, A., Maurizi, E., Barbero, F., Sala, M., Fattorini, S., Balletto, E., & Bonelli, S. (2015). 531
The Pied Piper: A Parasitic Beetle's Melodies Modulate Ant Behaviours. PLoS One, 532
10, e0130541. doi: 10.1371/journal.pone.0130541 533
Dinca, V., Montagud, S., Talavera, G., Hernandez-Roldan, J., Munguira, M. L., Garcia-534
Barros, E., Hebert, P.D.N, & Vila, R. (2015). DNA barcode reference library for 535
Page 22 of 31 Iberian butterflies enables a continental-scale preview of potential cryptic diversity. 536
Scientific Reports, 5. 537
Dinca, V., Zakharov, E. V., Hebert, P. D. N., & Vila, R. (2011). Complete DNA barcode 538
reference library for a country's butterfly fauna reveals high performance for 539
temperate Europe. Proceedings of the Royal Society B-Biological Sciences, 278, 540
347-355. 541
Donisthorpe, H. S. J. K. (1927). The Guests of British Ants. London: Routledge. 542
Downey, J. C., & Allyn, A. C. (1973). Butterfly ultrastructure: 1. Sound production and 543
associated abdominal structures in pupae of Lycaenidae and Riodinidae. Bulletin of 544
the Allyn Museum, 14, 1-47. 545
Downey, J. C., & Allyn, A. C. (1978). Sounds produced in pupae of Lycaenidae. Bulletin of 546
the Allyn Museum, 48, 1-14. 547
Elmes, G. W. (1989). The effect of multiple queens in small groups of Myrmica rubra L. 548
Actes colloquia. Insectes Sociaux, 5, 137 - 144. 549
Elmes, G. W., Akino, T., Thomas, J. A., Clarke, R. T., & Knapp, J. J. (2002). Interspecific 550
differences in cuticular hydrocarbon profiles of Myrmica ants are sufficiently 551
consistent to explain host specificity by Maculinea (large blue) butterflies. Oecologia, 552
130, 525-535. 553
Ferreira, R. S., Cros, E., Fresneau, D., & Rybak, F. (2014). Behavioural Contexts of Sound 554
Production in Pachycondyla Ants (Formicidae: Ponerinae). Acta Acustica united with 555
Acustica, 100, 739-747. 556
Ferreira, R. S., Poteaux, C., Delabie, J. H., Fresneau, D., & Rybak, F. (2010). Stridulations 557
reveal cryptic speciation in neotropical sympatric ants. PLoS One, 5, e15363. 558
Fiedler, K. (1991). Systematic, evolutionary, and ecological implications of myrmecophily 559
within the Lycaenidae (Insecta: Lepidoptera: Papilionoidea). Bonner Zoologische
560
Monographien, 31, 1-210
Page 23 of 31 Fiedler, K. (1992). Notes on the biology of Hypo/yeaena otbona (Lepidoptera: Lycaenidae) In 562
West Malaysia. Nachrichten des entomologischen Vereins Apollo, Frankfurt. NF, 13, 563
65-92. 564
Fiedler, K. (1994). Lycaenid butterflies and plants: is myrmecophily associated with amplified 565
hostplant diversity? Ecological Entomology, 19, 79-82. 566
Fiedler, K. (1998). Geographical patterns in life-history traits of Lycaenidae butterflies - 567
ecological and evolutionary patterns. Zoology, 100, 336 - 347. 568
Fiedler, K., Seufert, P., Maschwitz, U., & Idris, A. H. J. (1995). Notes on larval biology and 569
pupal morphology of Malaysian Curetis butterflies(Lepidoptera: Lycaenidae). 570
Transactions of the Lepidopterological Society of Japan, 45, 287-299. 571
Gerhardt, H. C., & Huber, F. (2002). Acoustic communication in insects and anurans: 572
common problems and diverse solutions: University of Chicago Press. 573
Gerrish, A. R. (1994). The influence of relatedness and resource investment on the 574
behaviour of worker ants towards brood. PhD, University of Exeter, Exeter, UK. 575
Golden, T. M. J., & Hill, P. S. (2016). The evolution of stridulatory communication in ants, 576
revisited. Insectes Sociaux, 63, 309-319. 577
Grafen, A. (1989). The Phylogenetic Regression. Philosophical Transactions of the Royal 578
Society of London Series B-Biological Sciences, 326, 119-157. doi: 579
10.1098/rstb.1989.0106 580
Gronenberg, W. (1996). Neuroethology of ants. Naturwissenschaften, 83, 15-27. 581
Hickling, R., & Brown, R. L. (2000). Analysis of acoustic communication by ants. Journal of 582
the Acoustical Society of America, 108, 1920-1929. 583
Hickling, R., & Brown, R. L. (2001). Response to "Ants ar deaf". Journal of the Acoustical 584
Society of America, 109, 3083. 585
Hill, C. J. (1993). The myrmecophilous organs of Arhopala madytus Fruhstorfer 586
(Lepidoptera: Lycaenidae). Australian Journal of Entomology, 32, 283-288. 587
Hill, P. S. (2009). How do animals use substrate-borne vibrations as an information source? 588
Naturwissenschaften, 96, 1355-1371. 589
Page 24 of 31 Hinton, H. E. (1951). Myrmecophilous Lycaenidae and other Lepidoptera - a summary. 590
Proceedings & Transactions of the South London Entomological and Natural History 591
Society, 1949-50, 111 - 175. 592
Hölldobler, B., Braun, U., Gronenberg, W., Kirchner, W. H., & Peeters, C. (1994). Trail 593
Communication in the Ant Megaponera Foetens (Fabr) (Formicidae, Ponerinae). 594
Journal of Insect Physiology, 40, 585-593. 595
Hölldobler, B., & Wilson, E. O. (1990). The ants. Berlin Heidelberg: Springer Verlag. 596
Hölldobler, B., & Wilson, E. O. (2009). The superorganism: the beauty, elegance, and 597
strangeness of insect societies. London: WW Norton & Company. 598
Howard, R. W. (1993). Cuticular Hydrocarbons and Chemical Communication. In D. W. 599
Stanley-Samuelson & D. R. Nelson (Eds.), Insect Lipids: Chemistry, Biochemistry 600
and Biology (pp. 179 - 226). Lincoln: University of Nebraska Press. 601
Hunt, J. H., & Richard, F.-J. (2013). Intracolony vibroacoustic communication in social 602
insects. Insectes Sociaux, 60, 403-417. 603
Janssen, E., Bestmann, H. J., Hölldobler, B., & Kern, F. (1995). N,N-dimethyluracil and 604
actinidine, two pheromones of the ponerine and Megaponera foetens (Fab.) 605
(Hymenoptera: Formicidae). Journal of Chemical Ecology, 21, 1947 - 1955. 606
Kirchner W. H. (1997) Acoustical communication in social insects. In: M. Lehrer (ed)
607
Orientation and communication in arthropods (pp. 273 – 300). Basel: Birkhäuser
608
Lenoir, A., D'Ettorre, P., Errard, C., & Hefetz, A. (2001). Chemical ecology and social 609
parasitism in ants. Annual Review of Entomology, 46, 573-599. 610
Malicky, H. (1969). Versuch einer Analyse der ökologischen Beziehungen zwischen 611
Lycaeniden (Lepidoptera) und Formiciden (Hymenoptera). Tijdschrift voor 612
Entomologie, 112, 213 - 298. 613
Markl, H. (1965). Stridulation in Leaf-Cutting ants. Science, 149, 1392 - 1393. 614
Markl, H. (1967). Die Verständigung durch Stridulationssignale bei Blattschneiderameisen: I 615
Die biologische Bedeutung der Stridulation. Zeitschrift für vergleichende Physiologie, 616
57, 299 - 330. 617
Page 25 of 31 Markl, H. (1973, September). The evolution of stridulatory communication in ants.
618
In Proceedings of the International Congress IUSSI, London (Vol. 7, pp. 258-265). 619
Martin, S. J., & Drijfhout, F. P. (2009). A review of ant cuticular hydrocarbons. Journal of 620
Chemical Ecology, 35, 1151-1161. 621
Menzel, J. G., & Tautz, J. (1994). Functional morphology of the subgenual organ of the 622
carpenter ant. Tissue and Cell, 26, 735 - 746. 623
Nash, D. R., & Boomsma, J. J. (2008). Communication between hosts and social parasites. 624
In P. d'Ettore & D. P. Hughes (Eds.), Sociobiology of Communication (pp. 55 - 79). 625
Oxford: Oxford University Press. 626
Pech, P., Fric, Z., Konvicka, M., & Zrzavy, J. (2004). Phylogeny of Maculinea blues 627
(Lepidoptera : Lycaenidae) based on morphological and ecological characters: 628
evolution of parasitic myrmecophily. Cladistics-the International Journal of the Willi 629
Hennig Society, 20, 362-375. 630
Pellissier, L., Litsios, G., Guisan, A., & Alvarez, N. (2012). Molecular substitution rate 631
increases in myrmecophilous lycaenid butterflies (Lepidoptera). Zoologica Scripta, 632
41, 651-658 633
Pierce, N. E., Braby, M. F., Heath, A., Lohman, D. J., Mathew, J., Rand, D. B., & Travassos, 634
M. A. (2002). The ecology and evolution of ant association in the Lycaenidae 635
(Lepidoptera). Annual Review of Entomology, 47, 733-771. 636
Pierce, N. E., Kitching, R. L., Buckley, R. C., Taylor, M. F. J., & Benbow, K. F. (1987). The 637
Costs and Benefits of Cooperation between the Australian Lycaenid Butterfly, 638
Jalmenus evagoras, and Its Attendant Ants. Behavioral Ecology and Sociobiology, 639
21, 237-248. 640
R Core Team. (2015). R: A Language and Environment for Statistical Computing. Vienna, 641
Austria: R Foundation for Statistical Computing. Retrieved from https://www.R-642
project.org 643
Riva, F., Barbero, F., Bonelli, S., Balletto, E., & Casacci, L. P. (in press). The acoustic 644
repertoire of lycaenid butterfly larvae. Bioacoustics. 645
Page 26 of 31 Roces, F., & Hölldobler, B. (1995). Vibrational communication between hitchhikers and 646
foragers in leaf-cutting ants (Atta cephalotes). Behavioral Ecology and Sociobiology, 647
37, 297-302. 648
Roces, F., & Hölldobler, B. (1996). Use of stridulation in foraging leaf cutting ants: 649
Mechanical support during cutting or short range recruitment signal? Behavioral 650
Ecology and Sociobiology, 39, 293-299. 651
Roces, F., & Tautz, J. (2001). Ants are deaf. Journal of the Acoustical Society of America, 652
109, 3080-3083. 653
Roces, F., Tautz, J., & Hölldobler, B. (1993). Stridulations in leaf-cutting ants. 654
Naturwissenschaften, 80, 521 - 524. 655
Ross, G. N. (1966). Life history studies on Mexican butterflies. IV. The ecology and ethology 656
of Anatole rossi, a myrmecophilous metalmark (Lepidoptera: Riodinidae). Annals of 657
the Entomological Society of America, 59, 985 - 1004. 658
Ruiz, E., Martinez, M. H., Martinez, M. D., & Hernandez, J. M. (2006). Morphological study of 659
the Stridulatory Organ in two species of Crematogaster genus: Crematogaster 660
scutellaris (Olivier 1792) and Crematogaster auberti (Emery 1869) (Hymenoptera : 661
Formicidae). Annales De La Societe Entomologique De France, 42, 99-105. 662
Sala, M., Casacci, L. P., Balletto, E., Bonelli, S., & Barbero, F. (2014). Variation in butterfly 663
larval acoustics as a strategy to infiltrate and exploit host ant colony resources. PLoS 664
One, 9, e94341. doi: 10.1371/journal.pone.0094341 665
Schurian, K. G., & Fiedler, K. (1991). Einfache Methoden zur Schallwahrnehmung bei 666
Bläulings-Larven (Lepidoptera: Lycaenidae). Entomologische Zeitschrift, 101, 393-667
412. 668
Tautz, J., Roces, F., & Hölldobler, B. (1995). Use of a sound based vibratome by leaf-cutting 669
ants. Science, 267, 84 - 87. 670
Thomas, J. A., Elmes, G. W., Schönrogge, K., Simcox, D. J., & Settele, J. (2005). Primary 671
hosts, secondary hosts and 'non-hosts': common confusions in the interpretation of 672
host specificity in Maculinea butterflies and other social parasites of ants. In J. 673
Page 27 of 31 Settele, E. Kühn & J. A. Thomas (Eds.), Studies on the ecology and conservation of 674
butterflies in Europe, vol. 2. (pp. 99 – 104). Sofia: Pensoft. 675
Thomas, J. A., Elmes, G. W., Sielezniew, M., Stankiewicz-Fiedurek, A., Simcox, D. J., 676
Settele, J., & Schonrogge, K. (2013). Mimetic host shifts in an endangered social 677
parasite of ants. Proceedings of the Royal Society B-Biological Sciences, 280, 678
20122336. http://dx.doi.org/10.1098/rspb.2012.2336 . 679
Thomas, J. A., Schönrogge, K., Bonelli, S., Barbero, F., & Balletto, E. (2010). Corruption of 680
ant acoustical signals by mimetic social parasites: Maculinea butterflies achieve 681
elevated status in host societies by mimicking the acoustics of queen ants. 682
Communicative & Integrative Biology, 3, 169-171. 683
Thomas, J. A., Schönrogge, K., & Elmes, G. W. (2005). Specializations and Host 684
Associations of Social Parasites of Ants. In M. D. E. Fellowes, G. J. Holloway & J. 685
Rolff (Eds.), Insect Evolutionary Ecology (pp. 475 - 514). London: Royal 686
Entomological Society. 687
Travassos, M. A., DeVries, P. J., & Pierce, N. E. (2008). A novel organ and mechanism for 688
larval sound production in butterfly caterpillars: Eurybia elvina. Tropical Lepidoptera 689
18, 20-23. 690
Travassos, M. A., & Pierce, N. E. (2000). Acoustics, context and function of vibrational 691
signalling in a lycaenid butterfly-ant mutualism. Animal Behaviour, 60, 13-26. 692
van Wilgenburg, E., Symonds, M. R., & Elgar, M. A. (2011). Evolution of cuticular 693
hydrocarbon diversity in ants. Journal of Evolutionary Biology, 24, 1188-1198. 694
vander Meer, R. K., & Morel, L. (1998). Nestmate Recognition in Ants. In R. K. vander Meer, 695
M. D. Breed, M. L. Winston & K. E. Espelie (Eds.), Pheromone Communication in 696
Social Insects (pp. 79 - 103). Oxford: Westview Press. 697
von Thienen, W., Metzler, D., Choe, D.-H., & Witte, V. (2014). Pheromone communication in 698
ants: a detailed analysis of concentration-dependent decisions in three species. 699
Behavioral Ecology and Sociobiology, 68, 1611-1627. 700
Wasmann, E. (1913). The ants and their guests. Smithonian Rep, 1912, 455 - 474. 701
Page 28 of 31 Wheeler, W. M. (1910). Ants:their structure, development and behavior (Vol. Vol. 9). New 702
York: Columbia University Press. 703
Wild, A. L. (2005). Taxonomic revision of the Pachycondyla apicalis species complex 704
(Hymenoptera: Formicidae). Zootaxa, 834, 1 - 25. 705
Winston, M. L. (1992). Semiochemicals and insect sociality. In B. D. Roitberg & M. B. Isman 706
(Eds.), Insect Chemical Ecology and Evolutionary Approach (pp. 315 - 333). London: 707
Chapman & Hall. 708
Witek, M., Barbero, F., & Marko, B. (2014). Myrmica ants host highly diverse parasitic 709
communities: from social parasites to microbes. Insectes Sociaux, 61, 307-323. 710
Page 29 of 31 Figures
711
Figure 1. The comparative morphology of sound production organs in myrmecophiles and 712
host ants. (a-e) the riodinids Synargis gela and Thisbe irenea (Riodinidae); larva (f, g) and 713
pupa (h-j) of the obligate lycaenid social parasite Maculinea rebeli and its adult host ant 714
Myrmica schencki (k-o); the adult beetle Paususs favieri (p-t) and its host Pheidole pallidula 715
(u-y). (a) Frontal view of Synargis gela head showing typical position of the riodinid vibratory 716
papillae; (b) general view of Thisbe irenea anterior edge of segment T-1 showing a vibratory 717
papilla (arrow) and the surface of the epicranium where the vibratory papilla strikes; (c) detail 718
of the vibratory papilla showing the annulations on its shaft and the epicranial granulations; 719
(d) enlarged view of the epicranial granulation and vibratory papilla; (e) details showing two 720
sizes of epicranial granulations. (f) Position of (g) the presumed sound producing organ of 721
Maculinea rebeli caterpillars and of its pupa (h), formed by a stridulatory plate (pars stridens) 722
placed on the fifth abdominal segment and a file (plectrum) in the sixth abdominal segment. 723
(k,p,u) Respective positions of the stridulatory organs of Myrmica schencki, Paussus favieri 724
and Pheidole pallidula; the organs are composed of suboval pars stridens (l,q,v) with minute 725
ridges (m,r,w) and a plectrum (n, x) consisting of a medial cuticular prominence (t,y) that 726
originates from the posterior edge of the postpetiole in the two ant species or of a curved row 727
of small cuticular spines in P. favieri (s,t). (a, modified by De Vries 1991; b-e modified by 728
DeVries 1988; p-y modified by Di Giulio et al. 2015). 729
730
Figure 2. A diagram of the phylogeny (left) and the cluster analysis constructed from a matrix 731
of pairwise normalized Euclidean distances of the sound profiles from three caterpillars of 13 732
species of lycaenid. Symbols and values refer to the intensity of interaction of the lycaenid 733
species with their host ants (0 = none; 4 = social parasite), following Fiedler (1991). 734
735 736 737 738
Page 30 of 31 Figure 1: 739 740 741 742 743 744 745 746 747 748 749
Page 31 of 31 Figure 2:
750